# Musk Acknowledges Anthropic’s Fable 5 Beats xAI’s Grok 4.5

By Simon Yoon

Canonical URL: https://www.tokenpost.com/news/technology/24437
Published: 2026-09-26T14:47:07.000Z
Updated: 2026-09-26T14:47:07.000Z
Section: Technology

> Musk said most tasks do not require Fable-level capability as xAI positioned Grok 4.5 for coding, agentic tasks and knowledge work at $2 per million input tokens.

Elon Musk acknowledged that Anthropic’s Fable 5 is more capable than xAI’s Grok 4.5, while arguing that most tasks do not require the higher-end model’s performance.

“In fairness, Fable is definitely better than Grok 4.5, but most tasks don’t require Fable-level capability,” Musk wrote July 8.

The next day, Musk added, “I was clearly wrong about Anthropic. They are obviously currently the leader in AI.” The statement was Musk’s assessment of Anthropic’s position, rather than an independently established industry ranking.

xAI announced Grok 4.5 on July 16 as a model for coding, agentic tasks and knowledge work. It listed the model at $2 per million input tokens and $6 per million output tokens.

In xAI’s published SWE-Bench Pro results, Grok 4.5 scored 64.7%. Anthropic’s Opus 4.8 scored 69.2%, while Anthropic’s Fable model scored 80.4% on the same coding benchmark.

Anthropic released Fable 5 and Mythos 5 June 9, then made Fable 5 broadly available beginning July 1. Fable 5 is Anthropic’s general-use model from its Mythos-class lineup.

The announcements put Musk’s comments alongside a clear performance comparison. Grok 4.5 trailed Anthropic’s Fable and Opus 4.8 on the cited benchmark, while xAI emphasized coding, agentic tasks and knowledge work and disclosed its token prices.